Start with a question, not a campaign

Most small business owners approach online advertising in one of two ways. Either they hold back entirely, convinced it will burn money, or they go in hard with a budget they cannot comfortably lose and hope the orders roll in. Both are expensive. The middle path — a deliberately small, tightly controlled test — is the one that actually teaches you something.

A good test is not a timid version of a campaign. It is a campaign with a single job: to answer a question. Will this audience respond to this offer at a price that still leaves me a profit? Everything else — the polished creative, the retargeting, the second and third audience — waits until you have that answer.

Pick one audience and stick to it

The single biggest reason small tests fail is that they try to test too much at once. If you split a £300 budget across three audiences, four adverts and two platforms, every result is too small to mean anything. You will end up with three vague hunches instead of one solid fact.

Choose the narrowest audience you can describe in a sentence. Something like:

• Homeowners within eight miles of your workshop who need a repair, not a full replacement.
• New parents in your town searching for a specific service in the next fortnight.
• Office managers at small firms in your city who book regular maintenance.

Write that sentence down before you touch any settings. Then set your location radius, your device preference if it matters, and your schedule to match it. Resist the temptation to add a second audience "just in case" — that is what the next test is for.

Make the offer unmistakable

A vague offer is the quiet killer of small budgets. "Get a quote" and "contact us for more information" give people nothing to say yes to. You need a specific, low-risk next step with a clear price or a clear outcome.

Good small-budget offers tend to look like this:

• A fixed-price service with a date attached, such as booked before the end of the month.
• A free, genuinely useful first step with a time limit, such as a twenty-minute consultation.
• A bundle or first-visit price that lets someone try you without committing fully.

Send every click to a dedicated landing page, not your homepage. Match the headline on that page to the headline in your advert word for word. Strip out the menu, keep one button, and make your phone number tappable on mobile. If a visitor has to think about what to do next, you have lost them.

Set a budget you can afford to lose

Decide the total amount before you start, and treat it as a lesson fee rather than an investment with a guaranteed return. For most small UK businesses, somewhere between £150 and £400 spread over ten to fourteen days is enough to gather a meaningful signal without stinging.

Work backwards from the numbers you need. If clicks in your sector cost around £1.50, then £300 buys roughly 200 clicks. At a 5% conversion rate that is ten enquiries — enough to judge whether the offer works. At a 1% conversion rate it is two, which tells you something different but still useful.

Set a daily cap rather than a monthly one, keep the bidding simple, and give the test a start and end date in your diary. Without an end date, small tests drift quietly into indefinite spending.

Set up the plumbing before you spend a penny

There is nothing more frustrating than a decent campaign you cannot measure. Before the first advert goes live, sort out three things:

• Conversion tracking on your form submission, booking or call button, so you know which enquiries came from the adverts.
• A call log or a dedicated number, because many small businesses get more phone calls than form fills and never see them in the platform's reports.
• A simple spreadsheet with columns for date, spend, clicks, enquiries and sales. Ten minutes a day is plenty.

Check the search terms report after the first couple of days and add obvious irrelevant searches as negatives. Then leave the campaign alone for a week. Constant fiddling resets the learning and makes the results impossible to read.

Read the numbers, then decide

At the end of the test, sit down with your spreadsheet and be honest. The decision usually falls into one of four buckets:

• Cheap clicks, good enquiries. Scale gradually — increase the daily budget by 20–30% each week and watch whether the cost per lead holds.
• Cheap clicks, no enquiries. Your audience is probably fine; the offer or the landing page is the weak link. Test one new offer next.
• Expensive clicks, few enquiries. The audience or keywords are too broad or too competitive. Narrow the targeting and try again.
• Nothing at all. That is a real result. Either the demand is not there, or the channel does not suit your business. Move the budget elsewhere with your head held high.

Change one variable at a time, keep your notes, and build on what you learn. A £300 test that saves you from spending £3,000 on the wrong campaign is one of the best investments your business will make this year.